Holographic memory systems involve the three-dimensional storage of holographic representations (i.e. holograms) of data elements as a pattern of varying refractive index and/or absorption imprinted in the volume of a storage medium such as a crystal of lithium niobate. Holographic memory systems are characterized by their high density storage potential and the potential speed with which the stored data is randomly accessed and transferred.
In general, holographic storage memory systems operate by combining a data-encoded signal beam with a reference beam to create an interference pattern throughout a photosensitive storage medium. The interference pattern induces material alterations in the storage medium that generate a hologram. The formation of the hologram in the storage medium is a function of the relative amplitudes and polarization states of, and phase differences between, the signal beam and the reference beam. It is also highly dependent on the incidents beam's wavelengths and angles at which the signal beam and the reference beam are projected into the storage medium.
Holographically stored data is reconstructed by projecting a reference beam similar to the reference beam used in storing the data into the storage medium at the same angle, wavelength, phase and position used to produce the hologram. The hologram and the referenced beam interact to reconstruct the signal beam. The reconstructed signal beam then is detected by using, for example, a photo-detector array. The recovered data then is post processed for delivery to output devices.
Various holographic storage drives have been designed in which a holographic medium in the form of a disk, card or the like, is used to record data and read the recorded data back in a manner similar to that of a magnetic hard disk drive used in personal computers. Such drives are typically provided with a mechanism to allow recording and access of holograms substantially throughout the spatial extent of the holographic disk or card. This allows such storage devices to advantageously store a relatively large amount of data. One such holographic drive was proposed by Tamarack Storage Devices, Inc. and is discussed in H. J. Coufal et al., Holographic Data Storage, Springer, pp. 343–357, which is hereby incorporated by reference.
FIG. 1 illustrates the general architecture of this drive. Beam splitter 12 splits a beam of coherent light emitted by laser 10 into a reference beam 14 and a signal beam 16. A spatial light modulator (“SLM”) and associated optics 18 encode signal beam 16 with data to be recorded in holographic media 20 and focus signal beam 16 into holographic media 20. After splitting, reference beam 14 is processed by phase or angle generating optics 22 and directed into transmissive holographic media 20 to form a hologram with signal beam 16. On reconstruction, the phase or angle of reference beam 14 is reproduced and the hologram regenerated in output beam 24. Output beam 24 is processed by CCD and associated optics 26 to generate electrical output data 28. All of the recording and reading optics and electronics are mounted in a head (not shown) which travels across the holographic media 20 to allow access to different locations in holographic media 20.
One drawback with the Tamarack design is that it uses a transmission geometry. That is, reconstruction takes place on the opposite side of holographic media 20 from signal beam 16. This means that reconstruction components such as CCD and associated optics 26 must be placed on the opposite side of holographic media 20 from the object beam components such as SLM and associated optics 18. As such, this type of architecture requires space on both sides of the holographic media which generally requires a relatively large volume. Thus, a holographic drive using such a transmission type architecture may be unsuitable for popular applications where relatively compact data storage is desirable, such as desktop or laptop computers.
A second holographic drive is disclosed in K. Saito and H. Horimai, Holographic 3-D Disk using In-line Face-to-Face Recording, Optical Media Laboratory, Sony Corporation (“Saito”). The general architecture of the device disclosed in Saito is shown in FIG. 2. During recording, a laser 50 projects coherent light through a collimation lens 52, into beam splitter 54 and towards SLM 56. A bitmapped pattern to be recorded is displayed in region (A) and region (R) is made transparent. In this way, light incident on region (R) generates a reference beam and light incident on region (A) generates the signal beam. The reference and signal beams then pass through objective lens 58 to reflective holographic recording medium 60 to record a hologram therein. As shown in FIG. 3, holographic medium 60 is made reflective by including a reflective surface 62 on a plane beneath a photosensitive holographic layer 64. In this way, a first hologram of the data input via incident signal beam 68a is formed in region (a) by reflected signal beam 68b interfering with incident reference beam 66a. A second hologram of the input data is formed in region (b) by incident signal beam 68a interfering with reflected reference beam 66b. 
Holographic media 60 is in the form of a disk which can be spatially translated to allow multiple holograms to be recorded therein with significant overlap between holograms. Thus, the holographic drive of Saito relies on shift multiplexing to store a relatively large number of holograms in holographic media 60.
Though not discussed in Saito, one method of effecting a readout of a hologram stored in medium 60 shown in FIG. 2 is to polarize region (R) of SLM 56 to block light emitted by laser 50 while region (A) remains transparent to such light. In this way, the reference beam to readout data stored in holographic medium 60 is provided by light passing through region (A) of SLM 56. After reflecting off of holographic medium 60, the reconstructed output beam can be polarized to pass through region (R) of SLM 56, through beam splitter 54 and to CCD and related optics 70.
As discussed above, the holographic drive disclosed in Saito uses reflective holographic media. As such, both the recording and readout optics and electronics can be placed on the same side of the media. In this way, the holographic drive of Saito overcomes a drawback of the Tamarack device of requiring a relatively large volume. However, in doing so, some other undesirable features are introduced. First, as noted above, the drive disclosed by Saito uses shift multiplexing. And, in a geometry in which the reference signal and output beams are normal to the recording surface (as disclosed by Saito) the selectivity between shift multiplexed holograms can be relatively low. That is, holograms that are at adjacent locations to a target hologram (e.g., a hologram that is being read) can generate interference when reading the target hologram. Such interference can lower the signal to noise ratio in the readout signal of the target holograms and ultimately lead to errors or severely limit the achievable density. This problem can be exacerbated by the fact that the signal and reference beams share the same angular bandwidth because beams near a center axis of the system will create holograms having very low selectivity.